A Comparative Study of CT Perfusion Postprocessing Tools in Medium/Distal Vessel Occlusion Stroke.

Abstract

Background and purpose: CTP has been validated for stroke due to large vessel occlusion, but not for medium or distal vessel occlusions (MDVO). The aim of this study was to evaluate discrepancy of 2 widely used postprocessing tools for CTP in patients with medium/distal vessel occlusion (MDVO) stroke.

Materials and methods: This retrospective single-center cohort study comprised CTP data sets of patients with acute isolated MDVO and disabling stroke. CTP postprocessing included Syngo.via and RapidAI. Perfusion lesions were segmented and compared with regard to volume and spatial distribution of the perfusion abnormalities. For each patient, the results of the 2 postprocessing tools were deemed to represent agreement regarding volume and/or distribution according to a predefined classification scheme. Penumbra volumes were compared between postprocessing tools by using the t test for paired samples.

Results: Fifty CTP data sets from 49 patients (mean age 75.3 ± 10.7 years, 33 male patients) were included. In <30% of patients, the predefined criteria for agreement were met which indicate at least 50% congruence regarding lesion size and lesion location as well as a maximum distance between lesion margins <50 mm. Comparison of tissue at risk volumes revealed outliers, volume disagreement, and distribution disagreement in up to 79.6%. The 2 postprocessing tools disagreed on presence and volume of an infarct core in up to 86% of patients. Penumbra volumes differed significantly between postprocessing tools (P < .001; mean difference 50.5 mL/16.9 mL).

Conclusions: Despite being in daily use, commercially available postprocessing tools for CTP provide discrepant results in patients with MDVO.